* [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
@ 2020-06-30 10:15 ` Russell King
0 siblings, 0 replies; 13+ messages in thread
From: Russell King @ 2020-06-30 10:15 UTC (permalink / raw)
To: Andrew Lunn, Florian Fainelli, Heiner Kallweit, René van Dorst
Cc: David S. Miller, netdev, Felix Fietkau, John Crispin, Sean Wang,
Mark Lee, Jakub Kicinski, Matthias Brugger, linux-arm-kernel,
linux-mediatek
The SGMII PCS PHY needs to be updated with the link configuration in
the mac_link_up() call rather than in mac_config(). However,
mtk_sgmii_setup_mode_force() programs the SGMII block during
mac_config() when using 802.3z interface modes with the link
configuration.
Split that functionality from mtk_sgmii_setup_mode_force(), moving it
to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
This does not look correct to me: 802.3z modes operate at a fixed
speed. The contents of mtk_sgmii_link_up() look more appropriate for
SGMII mode, but the original code definitely did not call
mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
René, can you assist with this patch please - I really think there are
problems with the existing code. You call mtk_sgmii_setup_mode_force()
in a block which is conditionalised as:
if (state->interface == PHY_INTERFACE_MODE_SGMII ||
phy_interface_mode_is_8023z(state->interface)) {
...
if (state->interface != PHY_INTERFACE_MODE_SGMII)
err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
state);
Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
2500BASE-X, which do not support anything but their native speeds.
Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
and 100M.
Note that this patch is more about moving uses of state->{speed,duplex}
into mac_link_up(), rather than fixing this problem, but I don't think
the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
use.
Thanks.
drivers/net/ethernet/mediatek/mtk_eth_soc.c | 9 ++++-
drivers/net/ethernet/mediatek/mtk_eth_soc.h | 3 +-
drivers/net/ethernet/mediatek/mtk_sgmii.c | 37 +++++++++++++++------
3 files changed, 36 insertions(+), 13 deletions(-)
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
index 20db302d31ce..ef9ec3b6a5c8 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
@@ -326,7 +326,7 @@ static void mtk_mac_config(struct phylink_config *config, unsigned int mode,
/* Setup SGMIISYS with the determined property */
if (state->interface != PHY_INTERFACE_MODE_SGMII)
err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
- state);
+ state->interface);
else if (phylink_autoneg_inband(mode))
err = mtk_sgmii_setup_mode_an(eth->sgmii, sid);
@@ -423,6 +423,13 @@ static void mtk_mac_link_up(struct phylink_config *config,
phylink_config);
u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
+ if (phy_interface_mode_is_8023z(interface)) {
+ /* Decide how GMAC and SGMIISYS be mapped */
+ int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
+ 0 : mac->id;
+ mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
+ }
+
mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
MAC_MCR_FORCE_RX_FC);
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
index 454cfcd465fd..6f4b99bb7bfb 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
@@ -932,7 +932,8 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct device_node *np,
u32 ana_rgc3);
int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id);
int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
- const struct phylink_link_state *state);
+ phy_interface_t interface);
+void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex);
void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id);
int mtk_gmac_sgmii_path_setup(struct mtk_eth *eth, int mac_id);
diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c b/drivers/net/ethernet/mediatek/mtk_sgmii.c
index 32d83421226a..372c85c830b5 100644
--- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
+++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
@@ -60,7 +60,7 @@ int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id)
}
int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
- const struct phylink_link_state *state)
+ phy_interface_t interface)
{
unsigned int val;
@@ -69,7 +69,7 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
regmap_read(ss->regmap[id], ss->ana_rgc3, &val);
val &= ~RG_PHY_SPEED_MASK;
- if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
+ if (interface == PHY_INTERFACE_MODE_2500BASEX)
val |= RG_PHY_SPEED_3_125G;
regmap_write(ss->regmap[id], ss->ana_rgc3, val);
@@ -78,11 +78,33 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
val &= ~SGMII_AN_ENABLE;
regmap_write(ss->regmap[id], SGMSYS_PCS_CONTROL_1, val);
+ if (interface == PHY_INTERFACE_MODE_1000BASEX ||
+ interface == PHY_INTERFACE_MODE_2500BASEX) {
+ /* SGMII force mode setting */
+ regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
+ val &= ~SGMII_IF_MODE_MASK;
+ val |= SGMII_SPEED_1000;
+ val |= SGMII_DUPLEX_FULL;
+ regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
+ }
+
+ /* Release PHYA power down state */
+ regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
+ val &= ~SGMII_PHYA_PWD;
+ regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
+
+ return 0;
+}
+
+void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex)
+{
+ unsigned int val;
+
/* SGMII force mode setting */
regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
val &= ~SGMII_IF_MODE_MASK;
- switch (state->speed) {
+ switch (speed) {
case SPEED_10:
val |= SGMII_SPEED_10;
break;
@@ -95,17 +117,10 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
break;
}
- if (state->duplex == DUPLEX_FULL)
+ if (duplex == DUPLEX_FULL)
val |= SGMII_DUPLEX_FULL;
regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
-
- /* Release PHYA power down state */
- regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
- val &= ~SGMII_PHYA_PWD;
- regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
-
- return 0;
}
void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id)
--
2.20.1
^ permalink raw reply related [flat|nested] 13+ messages in thread
* [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
@ 2020-06-30 10:15 ` Russell King
0 siblings, 0 replies; 13+ messages in thread
From: Russell King @ 2020-06-30 10:15 UTC (permalink / raw)
To: Andrew Lunn, Florian Fainelli, Heiner Kallweit, René van Dorst
Cc: netdev, Sean Wang, David S. Miller, linux-mediatek, John Crispin,
Matthias Brugger, Jakub Kicinski, Mark Lee, linux-arm-kernel,
Felix Fietkau
The SGMII PCS PHY needs to be updated with the link configuration in
the mac_link_up() call rather than in mac_config(). However,
mtk_sgmii_setup_mode_force() programs the SGMII block during
mac_config() when using 802.3z interface modes with the link
configuration.
Split that functionality from mtk_sgmii_setup_mode_force(), moving it
to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
This does not look correct to me: 802.3z modes operate at a fixed
speed. The contents of mtk_sgmii_link_up() look more appropriate for
SGMII mode, but the original code definitely did not call
mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
René, can you assist with this patch please - I really think there are
problems with the existing code. You call mtk_sgmii_setup_mode_force()
in a block which is conditionalised as:
if (state->interface == PHY_INTERFACE_MODE_SGMII ||
phy_interface_mode_is_8023z(state->interface)) {
...
if (state->interface != PHY_INTERFACE_MODE_SGMII)
err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
state);
Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
2500BASE-X, which do not support anything but their native speeds.
Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
and 100M.
Note that this patch is more about moving uses of state->{speed,duplex}
into mac_link_up(), rather than fixing this problem, but I don't think
the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
use.
Thanks.
drivers/net/ethernet/mediatek/mtk_eth_soc.c | 9 ++++-
drivers/net/ethernet/mediatek/mtk_eth_soc.h | 3 +-
drivers/net/ethernet/mediatek/mtk_sgmii.c | 37 +++++++++++++++------
3 files changed, 36 insertions(+), 13 deletions(-)
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
index 20db302d31ce..ef9ec3b6a5c8 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
@@ -326,7 +326,7 @@ static void mtk_mac_config(struct phylink_config *config, unsigned int mode,
/* Setup SGMIISYS with the determined property */
if (state->interface != PHY_INTERFACE_MODE_SGMII)
err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
- state);
+ state->interface);
else if (phylink_autoneg_inband(mode))
err = mtk_sgmii_setup_mode_an(eth->sgmii, sid);
@@ -423,6 +423,13 @@ static void mtk_mac_link_up(struct phylink_config *config,
phylink_config);
u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
+ if (phy_interface_mode_is_8023z(interface)) {
+ /* Decide how GMAC and SGMIISYS be mapped */
+ int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
+ 0 : mac->id;
+ mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
+ }
+
mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
MAC_MCR_FORCE_RX_FC);
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
index 454cfcd465fd..6f4b99bb7bfb 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
@@ -932,7 +932,8 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct device_node *np,
u32 ana_rgc3);
int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id);
int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
- const struct phylink_link_state *state);
+ phy_interface_t interface);
+void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex);
void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id);
int mtk_gmac_sgmii_path_setup(struct mtk_eth *eth, int mac_id);
diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c b/drivers/net/ethernet/mediatek/mtk_sgmii.c
index 32d83421226a..372c85c830b5 100644
--- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
+++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
@@ -60,7 +60,7 @@ int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id)
}
int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
- const struct phylink_link_state *state)
+ phy_interface_t interface)
{
unsigned int val;
@@ -69,7 +69,7 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
regmap_read(ss->regmap[id], ss->ana_rgc3, &val);
val &= ~RG_PHY_SPEED_MASK;
- if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
+ if (interface == PHY_INTERFACE_MODE_2500BASEX)
val |= RG_PHY_SPEED_3_125G;
regmap_write(ss->regmap[id], ss->ana_rgc3, val);
@@ -78,11 +78,33 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
val &= ~SGMII_AN_ENABLE;
regmap_write(ss->regmap[id], SGMSYS_PCS_CONTROL_1, val);
+ if (interface == PHY_INTERFACE_MODE_1000BASEX ||
+ interface == PHY_INTERFACE_MODE_2500BASEX) {
+ /* SGMII force mode setting */
+ regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
+ val &= ~SGMII_IF_MODE_MASK;
+ val |= SGMII_SPEED_1000;
+ val |= SGMII_DUPLEX_FULL;
+ regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
+ }
+
+ /* Release PHYA power down state */
+ regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
+ val &= ~SGMII_PHYA_PWD;
+ regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
+
+ return 0;
+}
+
+void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex)
+{
+ unsigned int val;
+
/* SGMII force mode setting */
regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
val &= ~SGMII_IF_MODE_MASK;
- switch (state->speed) {
+ switch (speed) {
case SPEED_10:
val |= SGMII_SPEED_10;
break;
@@ -95,17 +117,10 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
break;
}
- if (state->duplex == DUPLEX_FULL)
+ if (duplex == DUPLEX_FULL)
val |= SGMII_DUPLEX_FULL;
regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
-
- /* Release PHYA power down state */
- regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
- val &= ~SGMII_PHYA_PWD;
- regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
-
- return 0;
}
void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id)
--
2.20.1
_______________________________________________
Linux-mediatek mailing list
Linux-mediatek@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-mediatek
^ permalink raw reply related [flat|nested] 13+ messages in thread
* [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
@ 2020-06-30 10:15 ` Russell King
0 siblings, 0 replies; 13+ messages in thread
From: Russell King @ 2020-06-30 10:15 UTC (permalink / raw)
To: Andrew Lunn, Florian Fainelli, Heiner Kallweit, René van Dorst
Cc: netdev, Sean Wang, David S. Miller, linux-mediatek, John Crispin,
Matthias Brugger, Jakub Kicinski, Mark Lee, linux-arm-kernel,
Felix Fietkau
The SGMII PCS PHY needs to be updated with the link configuration in
the mac_link_up() call rather than in mac_config(). However,
mtk_sgmii_setup_mode_force() programs the SGMII block during
mac_config() when using 802.3z interface modes with the link
configuration.
Split that functionality from mtk_sgmii_setup_mode_force(), moving it
to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
This does not look correct to me: 802.3z modes operate at a fixed
speed. The contents of mtk_sgmii_link_up() look more appropriate for
SGMII mode, but the original code definitely did not call
mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
---
René, can you assist with this patch please - I really think there are
problems with the existing code. You call mtk_sgmii_setup_mode_force()
in a block which is conditionalised as:
if (state->interface == PHY_INTERFACE_MODE_SGMII ||
phy_interface_mode_is_8023z(state->interface)) {
...
if (state->interface != PHY_INTERFACE_MODE_SGMII)
err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
state);
Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
2500BASE-X, which do not support anything but their native speeds.
Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
and 100M.
Note that this patch is more about moving uses of state->{speed,duplex}
into mac_link_up(), rather than fixing this problem, but I don't think
the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
use.
Thanks.
drivers/net/ethernet/mediatek/mtk_eth_soc.c | 9 ++++-
drivers/net/ethernet/mediatek/mtk_eth_soc.h | 3 +-
drivers/net/ethernet/mediatek/mtk_sgmii.c | 37 +++++++++++++++------
3 files changed, 36 insertions(+), 13 deletions(-)
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
index 20db302d31ce..ef9ec3b6a5c8 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
@@ -326,7 +326,7 @@ static void mtk_mac_config(struct phylink_config *config, unsigned int mode,
/* Setup SGMIISYS with the determined property */
if (state->interface != PHY_INTERFACE_MODE_SGMII)
err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
- state);
+ state->interface);
else if (phylink_autoneg_inband(mode))
err = mtk_sgmii_setup_mode_an(eth->sgmii, sid);
@@ -423,6 +423,13 @@ static void mtk_mac_link_up(struct phylink_config *config,
phylink_config);
u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
+ if (phy_interface_mode_is_8023z(interface)) {
+ /* Decide how GMAC and SGMIISYS be mapped */
+ int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
+ 0 : mac->id;
+ mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
+ }
+
mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
MAC_MCR_FORCE_RX_FC);
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
index 454cfcd465fd..6f4b99bb7bfb 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
@@ -932,7 +932,8 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct device_node *np,
u32 ana_rgc3);
int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id);
int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
- const struct phylink_link_state *state);
+ phy_interface_t interface);
+void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex);
void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id);
int mtk_gmac_sgmii_path_setup(struct mtk_eth *eth, int mac_id);
diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c b/drivers/net/ethernet/mediatek/mtk_sgmii.c
index 32d83421226a..372c85c830b5 100644
--- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
+++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
@@ -60,7 +60,7 @@ int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id)
}
int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
- const struct phylink_link_state *state)
+ phy_interface_t interface)
{
unsigned int val;
@@ -69,7 +69,7 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
regmap_read(ss->regmap[id], ss->ana_rgc3, &val);
val &= ~RG_PHY_SPEED_MASK;
- if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
+ if (interface == PHY_INTERFACE_MODE_2500BASEX)
val |= RG_PHY_SPEED_3_125G;
regmap_write(ss->regmap[id], ss->ana_rgc3, val);
@@ -78,11 +78,33 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
val &= ~SGMII_AN_ENABLE;
regmap_write(ss->regmap[id], SGMSYS_PCS_CONTROL_1, val);
+ if (interface == PHY_INTERFACE_MODE_1000BASEX ||
+ interface == PHY_INTERFACE_MODE_2500BASEX) {
+ /* SGMII force mode setting */
+ regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
+ val &= ~SGMII_IF_MODE_MASK;
+ val |= SGMII_SPEED_1000;
+ val |= SGMII_DUPLEX_FULL;
+ regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
+ }
+
+ /* Release PHYA power down state */
+ regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
+ val &= ~SGMII_PHYA_PWD;
+ regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
+
+ return 0;
+}
+
+void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex)
+{
+ unsigned int val;
+
/* SGMII force mode setting */
regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
val &= ~SGMII_IF_MODE_MASK;
- switch (state->speed) {
+ switch (speed) {
case SPEED_10:
val |= SGMII_SPEED_10;
break;
@@ -95,17 +117,10 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
break;
}
- if (state->duplex == DUPLEX_FULL)
+ if (duplex == DUPLEX_FULL)
val |= SGMII_DUPLEX_FULL;
regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
-
- /* Release PHYA power down state */
- regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
- val &= ~SGMII_PHYA_PWD;
- regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
-
- return 0;
}
void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id)
--
2.20.1
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply related [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
2020-06-30 10:15 ` Russell King
(?)
@ 2020-06-30 10:46 ` Russell King - ARM Linux admin
-1 siblings, 0 replies; 13+ messages in thread
From: Russell King - ARM Linux admin @ 2020-06-30 10:46 UTC (permalink / raw)
To: Andrew Lunn, Florian Fainelli, Heiner Kallweit, René van Dorst
Cc: David S. Miller, netdev, Felix Fietkau, John Crispin, Sean Wang,
Mark Lee, Jakub Kicinski, Matthias Brugger, linux-arm-kernel,
linux-mediatek
On Tue, Jun 30, 2020 at 11:15:42AM +0100, Russell King wrote:
> The SGMII PCS PHY needs to be updated with the link configuration in
> the mac_link_up() call rather than in mac_config(). However,
> mtk_sgmii_setup_mode_force() programs the SGMII block during
> mac_config() when using 802.3z interface modes with the link
> configuration.
>
> Split that functionality from mtk_sgmii_setup_mode_force(), moving it
> to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
>
> This does not look correct to me: 802.3z modes operate at a fixed
> speed. The contents of mtk_sgmii_link_up() look more appropriate for
> SGMII mode, but the original code definitely did not call
> mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
>
> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
> ---
> René, can you assist with this patch please - I really think there are
> problems with the existing code. You call mtk_sgmii_setup_mode_force()
> in a block which is conditionalised as:
>
> if (state->interface == PHY_INTERFACE_MODE_SGMII ||
> phy_interface_mode_is_8023z(state->interface)) {
> ...
> if (state->interface != PHY_INTERFACE_MODE_SGMII)
> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
> state);
>
> Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
> 2500BASE-X, which do not support anything but their native speeds.
> Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
> and 100M.
>
> Note that this patch is more about moving uses of state->{speed,duplex}
> into mac_link_up(), rather than fixing this problem, but I don't think
> the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
> use.
My Coccinelle script just found this use of state->{speed,duplex} still
remaining:
if (MTK_HAS_CAPS(mac->hw->soc->caps,
MTK_TRGMII_MT7621_CLK)) {
...
} else {
if (state->interface !=
PHY_INTERFACE_MODE_TRGMII)
mtk_gmac0_rgmii_adjust(mac->hw,
state->speed);
which also needs to be eliminated. Can that also be moved to
mtk_mac_link_up()?
Thanks.
>
> Thanks.
>
> drivers/net/ethernet/mediatek/mtk_eth_soc.c | 9 ++++-
> drivers/net/ethernet/mediatek/mtk_eth_soc.h | 3 +-
> drivers/net/ethernet/mediatek/mtk_sgmii.c | 37 +++++++++++++++------
> 3 files changed, 36 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> index 20db302d31ce..ef9ec3b6a5c8 100644
> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> @@ -326,7 +326,7 @@ static void mtk_mac_config(struct phylink_config *config, unsigned int mode,
> /* Setup SGMIISYS with the determined property */
> if (state->interface != PHY_INTERFACE_MODE_SGMII)
> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
> - state);
> + state->interface);
> else if (phylink_autoneg_inband(mode))
> err = mtk_sgmii_setup_mode_an(eth->sgmii, sid);
>
> @@ -423,6 +423,13 @@ static void mtk_mac_link_up(struct phylink_config *config,
> phylink_config);
> u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
>
> + if (phy_interface_mode_is_8023z(interface)) {
> + /* Decide how GMAC and SGMIISYS be mapped */
> + int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
> + 0 : mac->id;
> + mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
> + }
> +
> mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
> MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
> MAC_MCR_FORCE_RX_FC);
> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
> index 454cfcd465fd..6f4b99bb7bfb 100644
> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
> @@ -932,7 +932,8 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct device_node *np,
> u32 ana_rgc3);
> int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id);
> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> - const struct phylink_link_state *state);
> + phy_interface_t interface);
> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex);
> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id);
>
> int mtk_gmac_sgmii_path_setup(struct mtk_eth *eth, int mac_id);
> diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c b/drivers/net/ethernet/mediatek/mtk_sgmii.c
> index 32d83421226a..372c85c830b5 100644
> --- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
> +++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
> @@ -60,7 +60,7 @@ int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id)
> }
>
> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> - const struct phylink_link_state *state)
> + phy_interface_t interface)
> {
> unsigned int val;
>
> @@ -69,7 +69,7 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
>
> regmap_read(ss->regmap[id], ss->ana_rgc3, &val);
> val &= ~RG_PHY_SPEED_MASK;
> - if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
> + if (interface == PHY_INTERFACE_MODE_2500BASEX)
> val |= RG_PHY_SPEED_3_125G;
> regmap_write(ss->regmap[id], ss->ana_rgc3, val);
>
> @@ -78,11 +78,33 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> val &= ~SGMII_AN_ENABLE;
> regmap_write(ss->regmap[id], SGMSYS_PCS_CONTROL_1, val);
>
> + if (interface == PHY_INTERFACE_MODE_1000BASEX ||
> + interface == PHY_INTERFACE_MODE_2500BASEX) {
> + /* SGMII force mode setting */
> + regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
> + val &= ~SGMII_IF_MODE_MASK;
> + val |= SGMII_SPEED_1000;
> + val |= SGMII_DUPLEX_FULL;
> + regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
> + }
> +
> + /* Release PHYA power down state */
> + regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
> + val &= ~SGMII_PHYA_PWD;
> + regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
> +
> + return 0;
> +}
> +
> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex)
> +{
> + unsigned int val;
> +
> /* SGMII force mode setting */
> regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
> val &= ~SGMII_IF_MODE_MASK;
>
> - switch (state->speed) {
> + switch (speed) {
> case SPEED_10:
> val |= SGMII_SPEED_10;
> break;
> @@ -95,17 +117,10 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> break;
> }
>
> - if (state->duplex == DUPLEX_FULL)
> + if (duplex == DUPLEX_FULL)
> val |= SGMII_DUPLEX_FULL;
>
> regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
> -
> - /* Release PHYA power down state */
> - regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
> - val &= ~SGMII_PHYA_PWD;
> - regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
> -
> - return 0;
> }
>
> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id)
> --
> 2.20.1
>
>
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
@ 2020-06-30 10:46 ` Russell King - ARM Linux admin
0 siblings, 0 replies; 13+ messages in thread
From: Russell King - ARM Linux admin @ 2020-06-30 10:46 UTC (permalink / raw)
To: Andrew Lunn, Florian Fainelli, Heiner Kallweit, René van Dorst
Cc: netdev, Sean Wang, David S. Miller, linux-mediatek, John Crispin,
Matthias Brugger, Jakub Kicinski, Mark Lee, linux-arm-kernel,
Felix Fietkau
On Tue, Jun 30, 2020 at 11:15:42AM +0100, Russell King wrote:
> The SGMII PCS PHY needs to be updated with the link configuration in
> the mac_link_up() call rather than in mac_config(). However,
> mtk_sgmii_setup_mode_force() programs the SGMII block during
> mac_config() when using 802.3z interface modes with the link
> configuration.
>
> Split that functionality from mtk_sgmii_setup_mode_force(), moving it
> to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
>
> This does not look correct to me: 802.3z modes operate at a fixed
> speed. The contents of mtk_sgmii_link_up() look more appropriate for
> SGMII mode, but the original code definitely did not call
> mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
>
> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
> ---
> René, can you assist with this patch please - I really think there are
> problems with the existing code. You call mtk_sgmii_setup_mode_force()
> in a block which is conditionalised as:
>
> if (state->interface == PHY_INTERFACE_MODE_SGMII ||
> phy_interface_mode_is_8023z(state->interface)) {
> ...
> if (state->interface != PHY_INTERFACE_MODE_SGMII)
> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
> state);
>
> Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
> 2500BASE-X, which do not support anything but their native speeds.
> Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
> and 100M.
>
> Note that this patch is more about moving uses of state->{speed,duplex}
> into mac_link_up(), rather than fixing this problem, but I don't think
> the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
> use.
My Coccinelle script just found this use of state->{speed,duplex} still
remaining:
if (MTK_HAS_CAPS(mac->hw->soc->caps,
MTK_TRGMII_MT7621_CLK)) {
...
} else {
if (state->interface !=
PHY_INTERFACE_MODE_TRGMII)
mtk_gmac0_rgmii_adjust(mac->hw,
state->speed);
which also needs to be eliminated. Can that also be moved to
mtk_mac_link_up()?
Thanks.
>
> Thanks.
>
> drivers/net/ethernet/mediatek/mtk_eth_soc.c | 9 ++++-
> drivers/net/ethernet/mediatek/mtk_eth_soc.h | 3 +-
> drivers/net/ethernet/mediatek/mtk_sgmii.c | 37 +++++++++++++++------
> 3 files changed, 36 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> index 20db302d31ce..ef9ec3b6a5c8 100644
> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> @@ -326,7 +326,7 @@ static void mtk_mac_config(struct phylink_config *config, unsigned int mode,
> /* Setup SGMIISYS with the determined property */
> if (state->interface != PHY_INTERFACE_MODE_SGMII)
> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
> - state);
> + state->interface);
> else if (phylink_autoneg_inband(mode))
> err = mtk_sgmii_setup_mode_an(eth->sgmii, sid);
>
> @@ -423,6 +423,13 @@ static void mtk_mac_link_up(struct phylink_config *config,
> phylink_config);
> u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
>
> + if (phy_interface_mode_is_8023z(interface)) {
> + /* Decide how GMAC and SGMIISYS be mapped */
> + int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
> + 0 : mac->id;
> + mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
> + }
> +
> mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
> MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
> MAC_MCR_FORCE_RX_FC);
> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
> index 454cfcd465fd..6f4b99bb7bfb 100644
> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
> @@ -932,7 +932,8 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct device_node *np,
> u32 ana_rgc3);
> int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id);
> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> - const struct phylink_link_state *state);
> + phy_interface_t interface);
> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex);
> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id);
>
> int mtk_gmac_sgmii_path_setup(struct mtk_eth *eth, int mac_id);
> diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c b/drivers/net/ethernet/mediatek/mtk_sgmii.c
> index 32d83421226a..372c85c830b5 100644
> --- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
> +++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
> @@ -60,7 +60,7 @@ int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id)
> }
>
> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> - const struct phylink_link_state *state)
> + phy_interface_t interface)
> {
> unsigned int val;
>
> @@ -69,7 +69,7 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
>
> regmap_read(ss->regmap[id], ss->ana_rgc3, &val);
> val &= ~RG_PHY_SPEED_MASK;
> - if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
> + if (interface == PHY_INTERFACE_MODE_2500BASEX)
> val |= RG_PHY_SPEED_3_125G;
> regmap_write(ss->regmap[id], ss->ana_rgc3, val);
>
> @@ -78,11 +78,33 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> val &= ~SGMII_AN_ENABLE;
> regmap_write(ss->regmap[id], SGMSYS_PCS_CONTROL_1, val);
>
> + if (interface == PHY_INTERFACE_MODE_1000BASEX ||
> + interface == PHY_INTERFACE_MODE_2500BASEX) {
> + /* SGMII force mode setting */
> + regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
> + val &= ~SGMII_IF_MODE_MASK;
> + val |= SGMII_SPEED_1000;
> + val |= SGMII_DUPLEX_FULL;
> + regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
> + }
> +
> + /* Release PHYA power down state */
> + regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
> + val &= ~SGMII_PHYA_PWD;
> + regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
> +
> + return 0;
> +}
> +
> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex)
> +{
> + unsigned int val;
> +
> /* SGMII force mode setting */
> regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
> val &= ~SGMII_IF_MODE_MASK;
>
> - switch (state->speed) {
> + switch (speed) {
> case SPEED_10:
> val |= SGMII_SPEED_10;
> break;
> @@ -95,17 +117,10 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> break;
> }
>
> - if (state->duplex == DUPLEX_FULL)
> + if (duplex == DUPLEX_FULL)
> val |= SGMII_DUPLEX_FULL;
>
> regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
> -
> - /* Release PHYA power down state */
> - regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
> - val &= ~SGMII_PHYA_PWD;
> - regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
> -
> - return 0;
> }
>
> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id)
> --
> 2.20.1
>
>
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
_______________________________________________
Linux-mediatek mailing list
Linux-mediatek@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-mediatek
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
@ 2020-06-30 10:46 ` Russell King - ARM Linux admin
0 siblings, 0 replies; 13+ messages in thread
From: Russell King - ARM Linux admin @ 2020-06-30 10:46 UTC (permalink / raw)
To: Andrew Lunn, Florian Fainelli, Heiner Kallweit, René van Dorst
Cc: netdev, Sean Wang, David S. Miller, linux-mediatek, John Crispin,
Matthias Brugger, Jakub Kicinski, Mark Lee, linux-arm-kernel,
Felix Fietkau
On Tue, Jun 30, 2020 at 11:15:42AM +0100, Russell King wrote:
> The SGMII PCS PHY needs to be updated with the link configuration in
> the mac_link_up() call rather than in mac_config(). However,
> mtk_sgmii_setup_mode_force() programs the SGMII block during
> mac_config() when using 802.3z interface modes with the link
> configuration.
>
> Split that functionality from mtk_sgmii_setup_mode_force(), moving it
> to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
>
> This does not look correct to me: 802.3z modes operate at a fixed
> speed. The contents of mtk_sgmii_link_up() look more appropriate for
> SGMII mode, but the original code definitely did not call
> mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
>
> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
> ---
> René, can you assist with this patch please - I really think there are
> problems with the existing code. You call mtk_sgmii_setup_mode_force()
> in a block which is conditionalised as:
>
> if (state->interface == PHY_INTERFACE_MODE_SGMII ||
> phy_interface_mode_is_8023z(state->interface)) {
> ...
> if (state->interface != PHY_INTERFACE_MODE_SGMII)
> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
> state);
>
> Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
> 2500BASE-X, which do not support anything but their native speeds.
> Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
> and 100M.
>
> Note that this patch is more about moving uses of state->{speed,duplex}
> into mac_link_up(), rather than fixing this problem, but I don't think
> the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
> use.
My Coccinelle script just found this use of state->{speed,duplex} still
remaining:
if (MTK_HAS_CAPS(mac->hw->soc->caps,
MTK_TRGMII_MT7621_CLK)) {
...
} else {
if (state->interface !=
PHY_INTERFACE_MODE_TRGMII)
mtk_gmac0_rgmii_adjust(mac->hw,
state->speed);
which also needs to be eliminated. Can that also be moved to
mtk_mac_link_up()?
Thanks.
>
> Thanks.
>
> drivers/net/ethernet/mediatek/mtk_eth_soc.c | 9 ++++-
> drivers/net/ethernet/mediatek/mtk_eth_soc.h | 3 +-
> drivers/net/ethernet/mediatek/mtk_sgmii.c | 37 +++++++++++++++------
> 3 files changed, 36 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> index 20db302d31ce..ef9ec3b6a5c8 100644
> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> @@ -326,7 +326,7 @@ static void mtk_mac_config(struct phylink_config *config, unsigned int mode,
> /* Setup SGMIISYS with the determined property */
> if (state->interface != PHY_INTERFACE_MODE_SGMII)
> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
> - state);
> + state->interface);
> else if (phylink_autoneg_inband(mode))
> err = mtk_sgmii_setup_mode_an(eth->sgmii, sid);
>
> @@ -423,6 +423,13 @@ static void mtk_mac_link_up(struct phylink_config *config,
> phylink_config);
> u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
>
> + if (phy_interface_mode_is_8023z(interface)) {
> + /* Decide how GMAC and SGMIISYS be mapped */
> + int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
> + 0 : mac->id;
> + mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
> + }
> +
> mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
> MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
> MAC_MCR_FORCE_RX_FC);
> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
> index 454cfcd465fd..6f4b99bb7bfb 100644
> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
> @@ -932,7 +932,8 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct device_node *np,
> u32 ana_rgc3);
> int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id);
> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> - const struct phylink_link_state *state);
> + phy_interface_t interface);
> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex);
> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id);
>
> int mtk_gmac_sgmii_path_setup(struct mtk_eth *eth, int mac_id);
> diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c b/drivers/net/ethernet/mediatek/mtk_sgmii.c
> index 32d83421226a..372c85c830b5 100644
> --- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
> +++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
> @@ -60,7 +60,7 @@ int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id)
> }
>
> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> - const struct phylink_link_state *state)
> + phy_interface_t interface)
> {
> unsigned int val;
>
> @@ -69,7 +69,7 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
>
> regmap_read(ss->regmap[id], ss->ana_rgc3, &val);
> val &= ~RG_PHY_SPEED_MASK;
> - if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
> + if (interface == PHY_INTERFACE_MODE_2500BASEX)
> val |= RG_PHY_SPEED_3_125G;
> regmap_write(ss->regmap[id], ss->ana_rgc3, val);
>
> @@ -78,11 +78,33 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> val &= ~SGMII_AN_ENABLE;
> regmap_write(ss->regmap[id], SGMSYS_PCS_CONTROL_1, val);
>
> + if (interface == PHY_INTERFACE_MODE_1000BASEX ||
> + interface == PHY_INTERFACE_MODE_2500BASEX) {
> + /* SGMII force mode setting */
> + regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
> + val &= ~SGMII_IF_MODE_MASK;
> + val |= SGMII_SPEED_1000;
> + val |= SGMII_DUPLEX_FULL;
> + regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
> + }
> +
> + /* Release PHYA power down state */
> + regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
> + val &= ~SGMII_PHYA_PWD;
> + regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
> +
> + return 0;
> +}
> +
> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex)
> +{
> + unsigned int val;
> +
> /* SGMII force mode setting */
> regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
> val &= ~SGMII_IF_MODE_MASK;
>
> - switch (state->speed) {
> + switch (speed) {
> case SPEED_10:
> val |= SGMII_SPEED_10;
> break;
> @@ -95,17 +117,10 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
> break;
> }
>
> - if (state->duplex == DUPLEX_FULL)
> + if (duplex == DUPLEX_FULL)
> val |= SGMII_DUPLEX_FULL;
>
> regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
> -
> - /* Release PHYA power down state */
> - regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
> - val &= ~SGMII_PHYA_PWD;
> - regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
> -
> - return 0;
> }
>
> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id)
> --
> 2.20.1
>
>
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
2020-06-30 10:15 ` Russell King
` (2 preceding siblings ...)
(?)
@ 2020-06-30 15:29 ` kernel test robot
-1 siblings, 0 replies; 13+ messages in thread
From: kernel test robot @ 2020-06-30 15:29 UTC (permalink / raw)
To: kbuild-all
[-- Attachment #1: Type: text/plain, Size: 26065 bytes --]
Hi Russell,
[FYI, it's a private test report for your RFC patch.]
[auto build test ERROR on net-next/master]
url: https://github.com/0day-ci/linux/commits/Russell-King/net-mtk_eth_soc-use-resolved-link-config-for-PCS-PHY/20200630-181720
base: https://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next.git 5fb62372a0207f1514fa6052c51991198c46ffe2
config: arm-allyesconfig (attached as .config)
compiler: arm-linux-gnueabi-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=arm
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>
All errors (new ones prefixed by >>):
drivers/net/phy/mdio-boardinfo.c:27: warning: Function parameter or member 'bus' not described in 'mdiobus_setup_mdiodev_from_board_info'
drivers/net/phy/mdio-boardinfo.c:27: warning: Function parameter or member 'cb' not described in 'mdiobus_setup_mdiodev_from_board_info'
drivers/net/phy/mdio-boardinfo.c:27: warning: Excess function parameter 'mdiodev' description in 'mdiobus_setup_mdiodev_from_board_info'
In file included from drivers/net/ethernet/mediatek/mtk_eth_soc.c:23:
drivers/net/ethernet/mediatek/mtk_eth_soc.c: In function 'mtk_mac_link_up':
>> drivers/net/ethernet/mediatek/mtk_eth_soc.c:428:27: error: 'eth' undeclared (first use in this function)
428 | int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
| ^~~
drivers/net/ethernet/mediatek/mtk_eth_soc.h:761:36: note: in definition of macro 'MTK_HAS_CAPS'
761 | #define MTK_HAS_CAPS(caps, _x) (((caps) & (_x)) == (_x))
| ^~~~
drivers/net/ethernet/mediatek/mtk_eth_soc.c:428:27: note: each undeclared identifier is reported only once for each function it appears in
428 | int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
| ^~~
drivers/net/ethernet/mediatek/mtk_eth_soc.h:761:36: note: in definition of macro 'MTK_HAS_CAPS'
761 | #define MTK_HAS_CAPS(caps, _x) (((caps) & (_x)) == (_x))
| ^~~~
make[4]: *** [scripts/Makefile.build:280: drivers/net/ethernet/mediatek/mtk_eth_soc.o] Error 1
drivers/net/wimax/i2400m/control.c:709: warning: Excess function parameter 'msg_skb' description in 'i2400m_msg_to_dev'
drivers/net/wimax/i2400m/control.c:1197: warning: Function parameter or member 'arg' not described in 'i2400m_set_init_config'
drivers/net/wimax/i2400m/control.c:1197: warning: Excess function parameter 'arg_size' description in 'i2400m_set_init_config'
In file included from include/linux/bits.h:23,
from include/linux/bitops.h:5,
from include/linux/kernel.h:12,
from include/linux/list.h:9,
from drivers/net/dsa/bcm_sf2_cfp.c:8:
drivers/net/dsa/bcm_sf2_cfp.c: In function 'udf_upper_bits':
include/linux/bits.h:26:28: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c:133:9: note: in expansion of macro 'GENMASK'
133 | return GENMASK(num_udf - 1, 0) >> (UDFS_PER_SLICE - 1);
| ^~~~~~~
include/linux/bits.h:26:40: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c:133:9: note: in expansion of macro 'GENMASK'
133 | return GENMASK(num_udf - 1, 0) >> (UDFS_PER_SLICE - 1);
| ^~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c: In function 'udf_lower_bits':
include/linux/bits.h:26:28: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c:138:13: note: in expansion of macro 'GENMASK'
138 | return (u8)GENMASK(num_udf - 1, 0);
| ^~~~~~~
include/linux/bits.h:26:40: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c:138:13: note: in expansion of macro 'GENMASK'
138 | return (u8)GENMASK(num_udf - 1, 0);
| ^~~~~~~
make[3]: *** [scripts/Makefile.build:497: drivers/net/ethernet/mediatek] Error 2
make[4]: Target '__build' not remade because of errors.
In file included from drivers/net/wireless/ath/carl9170/wlan.h:42,
from drivers/net/wireless/ath/carl9170/carl9170.h:57,
from drivers/net/wireless/ath/carl9170/main.c:47:
drivers/net/wireless/ath/carl9170/fwcmd.h:243:2: warning: alignment 1 of 'union <anonymous>' is less than 4 [-Wpacked-not-aligned]
243 | } __packed;
| ^
drivers/net/wireless/ath/carl9170/fwcmd.h:243:2: warning: alignment 1 of 'union <anonymous>' is less than 4 [-Wpacked-not-aligned]
In file included from drivers/net/wireless/ath/carl9170/carl9170.h:57,
from drivers/net/wireless/ath/carl9170/main.c:47:
drivers/net/wireless/ath/carl9170/wlan.h:373:1: warning: alignment 1 of 'struct ar9170_rx_frame_single' is less than 2 [-Wpacked-not-aligned]
373 | } __packed;
| ^
drivers/net/wireless/ath/carl9170/wlan.h:379:1: warning: alignment 1 of 'struct ar9170_rx_frame_head' is less than 2 [-Wpacked-not-aligned]
379 | } __packed;
| ^
drivers/net/wireless/ath/carl9170/wlan.h:384:1: warning: alignment 1 of 'struct ar9170_rx_frame_middle' is less than 2 [-Wpacked-not-aligned]
384 | } __packed;
| ^
drivers/net/wireless/ath/carl9170/wlan.h:390:1: warning: alignment 1 of 'struct ar9170_rx_frame_tail' is less than 2 [-Wpacked-not-aligned]
390 | } __packed;
| ^
drivers/net/bonding/bond_3ad.c:140: warning: Function parameter or member 'port' not described in '__get_first_agg'
drivers/net/bonding/bond_3ad.c:140: warning: Excess function parameter 'bond' description in '__get_first_agg'
drivers/net/bonding/bond_3ad.c:1655: warning: Function parameter or member 'agg' not described in 'ad_agg_selection_logic'
drivers/net/bonding/bond_3ad.c:1655: warning: Excess function parameter 'aggregator' description in 'ad_agg_selection_logic'
drivers/net/bonding/bond_3ad.c:1817: warning: Function parameter or member 'port' not described in 'ad_initialize_port'
drivers/net/bonding/bond_3ad.c:1817: warning: Excess function parameter 'aggregator' description in 'ad_initialize_port'
drivers/net/bonding/bond_3ad.c:1976: warning: Function parameter or member 'timeout' not described in 'bond_3ad_initiate_agg_selection'
--
drivers/net/phy/mdio-boardinfo.c:27: warning: Function parameter or member 'bus' not described in 'mdiobus_setup_mdiodev_from_board_info'
drivers/net/phy/mdio-boardinfo.c:27: warning: Function parameter or member 'cb' not described in 'mdiobus_setup_mdiodev_from_board_info'
drivers/net/phy/mdio-boardinfo.c:27: warning: Excess function parameter 'mdiodev' description in 'mdiobus_setup_mdiodev_from_board_info'
In file included from drivers/net/ethernet/mediatek/mtk_eth_soc.c:23:
drivers/net/ethernet/mediatek/mtk_eth_soc.c: In function 'mtk_mac_link_up':
>> drivers/net/ethernet/mediatek/mtk_eth_soc.c:428:27: error: 'eth' undeclared (first use in this function)
428 | int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
| ^~~
drivers/net/ethernet/mediatek/mtk_eth_soc.h:761:36: note: in definition of macro 'MTK_HAS_CAPS'
761 | #define MTK_HAS_CAPS(caps, _x) (((caps) & (_x)) == (_x))
| ^~~~
drivers/net/ethernet/mediatek/mtk_eth_soc.c:428:27: note: each undeclared identifier is reported only once for each function it appears in
428 | int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
| ^~~
drivers/net/ethernet/mediatek/mtk_eth_soc.h:761:36: note: in definition of macro 'MTK_HAS_CAPS'
761 | #define MTK_HAS_CAPS(caps, _x) (((caps) & (_x)) == (_x))
| ^~~~
make[4]: *** [scripts/Makefile.build:280: drivers/net/ethernet/mediatek/mtk_eth_soc.o] Error 1
drivers/net/wimax/i2400m/control.c:709: warning: Excess function parameter 'msg_skb' description in 'i2400m_msg_to_dev'
drivers/net/wimax/i2400m/control.c:1197: warning: Function parameter or member 'arg' not described in 'i2400m_set_init_config'
drivers/net/wimax/i2400m/control.c:1197: warning: Excess function parameter 'arg_size' description in 'i2400m_set_init_config'
In file included from include/linux/bits.h:23,
from include/linux/bitops.h:5,
from include/linux/kernel.h:12,
from include/linux/list.h:9,
from drivers/net/dsa/bcm_sf2_cfp.c:8:
drivers/net/dsa/bcm_sf2_cfp.c: In function 'udf_upper_bits':
include/linux/bits.h:26:28: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c:133:9: note: in expansion of macro 'GENMASK'
133 | return GENMASK(num_udf - 1, 0) >> (UDFS_PER_SLICE - 1);
| ^~~~~~~
include/linux/bits.h:26:40: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c:133:9: note: in expansion of macro 'GENMASK'
133 | return GENMASK(num_udf - 1, 0) >> (UDFS_PER_SLICE - 1);
| ^~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c: In function 'udf_lower_bits':
include/linux/bits.h:26:28: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c:138:13: note: in expansion of macro 'GENMASK'
138 | return (u8)GENMASK(num_udf - 1, 0);
| ^~~~~~~
include/linux/bits.h:26:40: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/dsa/bcm_sf2_cfp.c:138:13: note: in expansion of macro 'GENMASK'
138 | return (u8)GENMASK(num_udf - 1, 0);
| ^~~~~~~
make[3]: *** [scripts/Makefile.build:497: drivers/net/ethernet/mediatek] Error 2
make[4]: Target '__build' not remade because of errors.
In file included from drivers/net/wireless/ath/carl9170/wlan.h:42,
from drivers/net/wireless/ath/carl9170/carl9170.h:57,
from drivers/net/wireless/ath/carl9170/main.c:47:
drivers/net/wireless/ath/carl9170/fwcmd.h:243:2: warning: alignment 1 of 'union <anonymous>' is less than 4 [-Wpacked-not-aligned]
243 | } __packed;
| ^
drivers/net/wireless/ath/carl9170/fwcmd.h:243:2: warning: alignment 1 of 'union <anonymous>' is less than 4 [-Wpacked-not-aligned]
In file included from drivers/net/wireless/ath/carl9170/carl9170.h:57,
from drivers/net/wireless/ath/carl9170/main.c:47:
drivers/net/wireless/ath/carl9170/wlan.h:373:1: warning: alignment 1 of 'struct ar9170_rx_frame_single' is less than 2 [-Wpacked-not-aligned]
373 | } __packed;
| ^
drivers/net/wireless/ath/carl9170/wlan.h:379:1: warning: alignment 1 of 'struct ar9170_rx_frame_head' is less than 2 [-Wpacked-not-aligned]
379 | } __packed;
| ^
drivers/net/wireless/ath/carl9170/wlan.h:384:1: warning: alignment 1 of 'struct ar9170_rx_frame_middle' is less than 2 [-Wpacked-not-aligned]
384 | } __packed;
| ^
drivers/net/wireless/ath/carl9170/wlan.h:390:1: warning: alignment 1 of 'struct ar9170_rx_frame_tail' is less than 2 [-Wpacked-not-aligned]
390 | } __packed;
| ^
drivers/net/bonding/bond_3ad.c:140: warning: Function parameter or member 'port' not described in '__get_first_agg'
drivers/net/bonding/bond_3ad.c:140: warning: Excess function parameter 'bond' description in '__get_first_agg'
drivers/net/bonding/bond_3ad.c:1655: warning: Function parameter or member 'agg' not described in 'ad_agg_selection_logic'
drivers/net/bonding/bond_3ad.c:1655: warning: Excess function parameter 'aggregator' description in 'ad_agg_selection_logic'
drivers/net/bonding/bond_3ad.c:1817: warning: Function parameter or member 'port' not described in 'ad_initialize_port'
drivers/net/bonding/bond_3ad.c:1817: warning: Excess function parameter 'aggregator' description in 'ad_initialize_port'
drivers/net/bonding/bond_3ad.c:1976: warning: Function parameter or member 'timeout' not described in 'bond_3ad_initiate_agg_selection'
--
In file included from drivers/net/ethernet/mediatek/mtk_eth_soc.c:23:
drivers/net/ethernet/mediatek/mtk_eth_soc.c: In function 'mtk_mac_link_up':
>> drivers/net/ethernet/mediatek/mtk_eth_soc.c:428:27: error: 'eth' undeclared (first use in this function)
428 | int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
| ^~~
drivers/net/ethernet/mediatek/mtk_eth_soc.h:761:36: note: in definition of macro 'MTK_HAS_CAPS'
761 | #define MTK_HAS_CAPS(caps, _x) (((caps) & (_x)) == (_x))
| ^~~~
drivers/net/ethernet/mediatek/mtk_eth_soc.c:428:27: note: each undeclared identifier is reported only once for each function it appears in
428 | int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
| ^~~
drivers/net/ethernet/mediatek/mtk_eth_soc.h:761:36: note: in definition of macro 'MTK_HAS_CAPS'
761 | #define MTK_HAS_CAPS(caps, _x) (((caps) & (_x)) == (_x))
| ^~~~
make[4]: *** [scripts/Makefile.build:280: drivers/net/ethernet/mediatek/mtk_eth_soc.o] Error 1
make[3]: *** [scripts/Makefile.build:497: drivers/net/ethernet/mediatek] Error 2
make[4]: Target '__build' not remade because of errors.
drivers/net/ethernet/3com/3c574_cs.c: In function 'update_stats':
drivers/net/ethernet/3com/3c574_cs.c:954:9: warning: variable 'tx' set but not used [-Wunused-but-set-variable]
954 | u8 rx, tx, up;
| ^~
drivers/net/ethernet/3com/3c574_cs.c:954:5: warning: variable 'rx' set but not used [-Wunused-but-set-variable]
954 | u8 rx, tx, up;
| ^~
drivers/net/ethernet/agere/et131x.c: In function 'et1310_setup_device_for_multicast':
drivers/net/ethernet/agere/et131x.c:953:6: warning: variable 'pm_csr' set but not used [-Wunused-but-set-variable]
953 | u32 pm_csr;
| ^~~~~~
drivers/net/ethernet/agere/et131x.c: In function 'et1310_setup_device_for_unicast':
drivers/net/ethernet/agere/et131x.c:1002:6: warning: variable 'pm_csr' set but not used [-Wunused-but-set-variable]
1002 | u32 pm_csr;
| ^~~~~~
drivers/net/ethernet/agere/et131x.c: In function 'et131x_isr':
drivers/net/ethernet/agere/et131x.c:3446:8: warning: variable 'pm_csr' set but not used [-Wunused-but-set-variable]
3446 | u32 pm_csr;
| ^~~~~~
drivers/net/ethernet/8390/axnet_cs.c:907: warning: Function parameter or member 'txqueue' not described in 'axnet_tx_timeout'
In file included from include/linux/bits.h:23,
from include/linux/bitops.h:5,
from include/linux/bitmap.h:8,
from include/linux/ethtool.h:16,
from drivers/net/ethernet/apm/xgene/xgene_enet_ethtool.c:8:
drivers/net/ethernet/apm/xgene/xgene_enet_ethtool.c: In function 'xgene_get_extd_stats':
include/linux/bits.h:26:28: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/ethernet/apm/xgene/xgene_enet_ethtool.c:225:11: note: in expansion of macro 'GENMASK'
225 | mask = GENMASK(gstrings_extd_stats[i].mask - 1, 0);
| ^~~~~~~
include/linux/bits.h:26:40: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
26 | __builtin_constant_p((l) > (h)), (l) > (h), 0)))
| ^
include/linux/build_bug.h:16:62: note: in definition of macro 'BUILD_BUG_ON_ZERO'
16 | #define BUILD_BUG_ON_ZERO(e) ((int)(sizeof(struct { int:(-!!(e)); })))
| ^
include/linux/bits.h:39:3: note: in expansion of macro 'GENMASK_INPUT_CHECK'
39 | (GENMASK_INPUT_CHECK(h, l) + __GENMASK(h, l))
| ^~~~~~~~~~~~~~~~~~~
drivers/net/ethernet/apm/xgene/xgene_enet_ethtool.c:225:11: note: in expansion of macro 'GENMASK'
225 | mask = GENMASK(gstrings_extd_stats[i].mask - 1, 0);
| ^~~~~~~
drivers/net/ethernet/apm/xgene-v2/main.c:725:36: warning: 'xge_acpi_match' defined but not used [-Wunused-const-variable=]
725 | static const struct acpi_device_id xge_acpi_match[] = {
| ^~~~~~~~~~~~~~
In file included from drivers/net/ethernet/aquantia/atlantic/aq_pci_func.c:20:
drivers/net/ethernet/aquantia/atlantic/hw_atl2/hw_atl2.h:9:10: fatal error: aq_common.h: No such file or directory
9 | #include "aq_common.h"
| ^~~~~~~~~~~~~
compilation terminated.
make[5]: *** [scripts/Makefile.build:281: drivers/net/ethernet/aquantia/atlantic/aq_pci_func.o] Error 1
drivers/net/ethernet/aquantia/atlantic/aq_filters.c: In function 'aq_check_approve_fvlan':
drivers/net/ethernet/aquantia/atlantic/aq_filters.c:158:20: warning: comparison of unsigned expression < 0 is always false [-Wtype-limits]
158 | if (fsp->location < AQ_RX_FIRST_LOC_FVLANID ||
| ^
drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c:1516: warning: Cannot understand * @brief Set VLAN filter table
on line 1516 - I thought it was a doc line
drivers/net/ethernet/aquantia/atlantic/hw_atl2/hw_atl2.c:6:10: fatal error: aq_hw.h: No such file or directory
6 | #include "aq_hw.h"
| ^~~~~~~~~
compilation terminated.
make[5]: *** [scripts/Makefile.build:281: drivers/net/ethernet/aquantia/atlantic/hw_atl2/hw_atl2.o] Error 1
drivers/net/ethernet/aquantia/atlantic/hw_atl2/hw_atl2_utils.c:8:10: fatal error: aq_hw_utils.h: No such file or directory
8 | #include "aq_hw_utils.h"
| ^~~~~~~~~~~~~~~
compilation terminated.
make[5]: *** [scripts/Makefile.build:281: drivers/net/ethernet/aquantia/atlantic/hw_atl2/hw_atl2_utils.o] Error 1
drivers/net/ethernet/aquantia/atlantic/hw_atl2/hw_atl2_llh.c:8:10: fatal error: aq_hw_utils.h: No such file or directory
8 | #include "aq_hw_utils.h"
| ^~~~~~~~~~~~~~~
compilation terminated.
make[5]: *** [scripts/Makefile.build:281: drivers/net/ethernet/aquantia/atlantic/hw_atl2/hw_atl2_llh.o] Error 1
drivers/net/ethernet/aquantia/atlantic/hw_atl2/hw_atl2_utils_fw.c:8:10: fatal error: aq_hw.h: No such file or directory
8 | #include "aq_hw.h"
| ^~~~~~~~~
compilation terminated.
make[5]: *** [scripts/Makefile.build:281: drivers/net/ethernet/aquantia/atlantic/hw_atl2/hw_atl2_utils_fw.o] Error 1
In file included from drivers/net/ethernet/aquantia/atlantic/macsec/macsec_api.c:6:
..
vim +/eth +428 drivers/net/ethernet/mediatek/mtk_eth_soc.c
416
417 static void mtk_mac_link_up(struct phylink_config *config,
418 struct phy_device *phy,
419 unsigned int mode, phy_interface_t interface,
420 int speed, int duplex, bool tx_pause, bool rx_pause)
421 {
422 struct mtk_mac *mac = container_of(config, struct mtk_mac,
423 phylink_config);
424 u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
425
426 if (phy_interface_mode_is_8023z(interface)) {
427 /* Decide how GMAC and SGMIISYS be mapped */
> 428 int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
429 0 : mac->id;
430 mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
431 }
432
433 mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
434 MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
435 MAC_MCR_FORCE_RX_FC);
436
437 /* Configure speed */
438 switch (speed) {
439 case SPEED_2500:
440 case SPEED_1000:
441 mcr |= MAC_MCR_SPEED_1000;
442 break;
443 case SPEED_100:
444 mcr |= MAC_MCR_SPEED_100;
445 break;
446 }
447
448 /* Configure duplex */
449 if (duplex == DUPLEX_FULL)
450 mcr |= MAC_MCR_FORCE_DPX;
451
452 /* Configure pause modes - phylink will avoid these for half duplex */
453 if (tx_pause)
454 mcr |= MAC_MCR_FORCE_TX_FC;
455 if (rx_pause)
456 mcr |= MAC_MCR_FORCE_RX_FC;
457
458 mcr |= MAC_MCR_TX_EN | MAC_MCR_RX_EN;
459 mtk_w32(mac->hw, mcr, MTK_MAC_MCR(mac->id));
460 }
461
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org
[-- Attachment #2: config.gz --]
[-- Type: application/gzip, Size: 75682 bytes --]
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
2020-06-30 10:46 ` Russell King - ARM Linux admin
(?)
@ 2020-06-30 22:13 ` René van Dorst
-1 siblings, 0 replies; 13+ messages in thread
From: René van Dorst @ 2020-06-30 22:13 UTC (permalink / raw)
To: Russell King - ARM Linux admin, Sean Wang
Cc: Andrew Lunn, Florian Fainelli, Heiner Kallweit, David S. Miller,
netdev, Felix Fietkau, John Crispin, Mark Lee, Jakub Kicinski,
Matthias Brugger, linux-arm-kernel, linux-mediatek
Hi Russel and Sean,
Quoting Russell King - ARM Linux admin <linux@armlinux.org.uk>:
> On Tue, Jun 30, 2020 at 11:15:42AM +0100, Russell King wrote:
>> The SGMII PCS PHY needs to be updated with the link configuration in
>> the mac_link_up() call rather than in mac_config(). However,
>> mtk_sgmii_setup_mode_force() programs the SGMII block during
>> mac_config() when using 802.3z interface modes with the link
>> configuration.
>>
>> Split that functionality from mtk_sgmii_setup_mode_force(), moving it
>> to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
>>
>> This does not look correct to me: 802.3z modes operate at a fixed
>> speed. The contents of mtk_sgmii_link_up() look more appropriate for
>> SGMII mode, but the original code definitely did not call
>> mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
>>
>> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
>> ---
>> René, can you assist with this patch please - I really think there are
>> problems with the existing code. You call mtk_sgmii_setup_mode_force()
>> in a block which is conditionalised as:
>>
>> if (state->interface == PHY_INTERFACE_MODE_SGMII ||
>> phy_interface_mode_is_8023z(state->interface)) {
>> ...
>> if (state->interface != PHY_INTERFACE_MODE_SGMII)
>> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
>> state);
>>
>> Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
>> 2500BASE-X, which do not support anything but their native speeds.
>> Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
>> and 100M.
>>
>> Note that this patch is more about moving uses of state->{speed,duplex}
>> into mac_link_up(), rather than fixing this problem, but I don't think
>> the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
>> use.
>
> My Coccinelle script just found this use of state->{speed,duplex} still
> remaining:
>
> if (MTK_HAS_CAPS(mac->hw->soc->caps,
> MTK_TRGMII_MT7621_CLK)) {
> ...
> } else {
> if (state->interface !=
> PHY_INTERFACE_MODE_TRGMII)
> mtk_gmac0_rgmii_adjust(mac->hw,
> state->speed);
>
> which also needs to be eliminated. Can that also be moved to
> mtk_mac_link_up()?
I know, you have pointed that out before. But I don't know how to fix
mtk_gmac0_rgmii_adjust(). This function changes the PLL of the MAC.
But without
documentation I am not sure what all the bits are used for.
Begin April I had a conversation with Sean about this. I also
explained what the
issue was. AFAIK he was going to take care of this issue.
Sean did you had time to resolve this issue?
Greats,
René
>
> Thanks.
>
>>
>> Thanks.
>>
>> drivers/net/ethernet/mediatek/mtk_eth_soc.c | 9 ++++-
>> drivers/net/ethernet/mediatek/mtk_eth_soc.h | 3 +-
>> drivers/net/ethernet/mediatek/mtk_sgmii.c | 37 +++++++++++++++------
>> 3 files changed, 36 insertions(+), 13 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> index 20db302d31ce..ef9ec3b6a5c8 100644
>> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> @@ -326,7 +326,7 @@ static void mtk_mac_config(struct
>> phylink_config *config, unsigned int mode,
>> /* Setup SGMIISYS with the determined property */
>> if (state->interface != PHY_INTERFACE_MODE_SGMII)
>> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
>> - state);
>> + state->interface);
>> else if (phylink_autoneg_inband(mode))
>> err = mtk_sgmii_setup_mode_an(eth->sgmii, sid);
>>
>> @@ -423,6 +423,13 @@ static void mtk_mac_link_up(struct
>> phylink_config *config,
>> phylink_config);
>> u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
>>
>> + if (phy_interface_mode_is_8023z(interface)) {
>> + /* Decide how GMAC and SGMIISYS be mapped */
>> + int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
>> + 0 : mac->id;
>> + mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
>> + }
>> +
>> mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
>> MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
>> MAC_MCR_FORCE_RX_FC);
>> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> index 454cfcd465fd..6f4b99bb7bfb 100644
>> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> @@ -932,7 +932,8 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct
>> device_node *np,
>> u32 ana_rgc3);
>> int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id);
>> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
>> - const struct phylink_link_state *state);
>> + phy_interface_t interface);
>> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed,
>> int duplex);
>> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id);
>>
>> int mtk_gmac_sgmii_path_setup(struct mtk_eth *eth, int mac_id);
>> diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> b/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> index 32d83421226a..372c85c830b5 100644
>> --- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> +++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> @@ -60,7 +60,7 @@ int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id)
>> }
>>
>> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
>> - const struct phylink_link_state *state)
>> + phy_interface_t interface)
>> {
>> unsigned int val;
>>
>> @@ -69,7 +69,7 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii
>> *ss, int id,
>>
>> regmap_read(ss->regmap[id], ss->ana_rgc3, &val);
>> val &= ~RG_PHY_SPEED_MASK;
>> - if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
>> + if (interface == PHY_INTERFACE_MODE_2500BASEX)
>> val |= RG_PHY_SPEED_3_125G;
>> regmap_write(ss->regmap[id], ss->ana_rgc3, val);
>>
>> @@ -78,11 +78,33 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii
>> *ss, int id,
>> val &= ~SGMII_AN_ENABLE;
>> regmap_write(ss->regmap[id], SGMSYS_PCS_CONTROL_1, val);
>>
>> + if (interface == PHY_INTERFACE_MODE_1000BASEX ||
>> + interface == PHY_INTERFACE_MODE_2500BASEX) {
>> + /* SGMII force mode setting */
>> + regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
>> + val &= ~SGMII_IF_MODE_MASK;
>> + val |= SGMII_SPEED_1000;
>> + val |= SGMII_DUPLEX_FULL;
>> + regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
>> + }
>> +
>> + /* Release PHYA power down state */
>> + regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
>> + val &= ~SGMII_PHYA_PWD;
>> + regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
>> +
>> + return 0;
>> +}
>> +
>> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex)
>> +{
>> + unsigned int val;
>> +
>> /* SGMII force mode setting */
>> regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
>> val &= ~SGMII_IF_MODE_MASK;
>>
>> - switch (state->speed) {
>> + switch (speed) {
>> case SPEED_10:
>> val |= SGMII_SPEED_10;
>> break;
>> @@ -95,17 +117,10 @@ int mtk_sgmii_setup_mode_force(struct
>> mtk_sgmii *ss, int id,
>> break;
>> }
>>
>> - if (state->duplex == DUPLEX_FULL)
>> + if (duplex == DUPLEX_FULL)
>> val |= SGMII_DUPLEX_FULL;
>>
>> regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
>> -
>> - /* Release PHYA power down state */
>> - regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
>> - val &= ~SGMII_PHYA_PWD;
>> - regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
>> -
>> - return 0;
>> }
>>
>> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id)
>> --
>> 2.20.1
>>
>>
>
> --
> RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
> FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
@ 2020-06-30 22:13 ` René van Dorst
0 siblings, 0 replies; 13+ messages in thread
From: René van Dorst @ 2020-06-30 22:13 UTC (permalink / raw)
To: Russell King - ARM Linux admin, Sean Wang
Cc: Andrew Lunn, Florian Fainelli, netdev, Mark Lee, linux-mediatek,
John Crispin, Matthias Brugger, Jakub Kicinski, Heiner Kallweit,
David S. Miller, linux-arm-kernel, Felix Fietkau
Hi Russel and Sean,
Quoting Russell King - ARM Linux admin <linux@armlinux.org.uk>:
> On Tue, Jun 30, 2020 at 11:15:42AM +0100, Russell King wrote:
>> The SGMII PCS PHY needs to be updated with the link configuration in
>> the mac_link_up() call rather than in mac_config(). However,
>> mtk_sgmii_setup_mode_force() programs the SGMII block during
>> mac_config() when using 802.3z interface modes with the link
>> configuration.
>>
>> Split that functionality from mtk_sgmii_setup_mode_force(), moving it
>> to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
>>
>> This does not look correct to me: 802.3z modes operate at a fixed
>> speed. The contents of mtk_sgmii_link_up() look more appropriate for
>> SGMII mode, but the original code definitely did not call
>> mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
>>
>> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
>> ---
>> René, can you assist with this patch please - I really think there are
>> problems with the existing code. You call mtk_sgmii_setup_mode_force()
>> in a block which is conditionalised as:
>>
>> if (state->interface == PHY_INTERFACE_MODE_SGMII ||
>> phy_interface_mode_is_8023z(state->interface)) {
>> ...
>> if (state->interface != PHY_INTERFACE_MODE_SGMII)
>> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
>> state);
>>
>> Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
>> 2500BASE-X, which do not support anything but their native speeds.
>> Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
>> and 100M.
>>
>> Note that this patch is more about moving uses of state->{speed,duplex}
>> into mac_link_up(), rather than fixing this problem, but I don't think
>> the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
>> use.
>
> My Coccinelle script just found this use of state->{speed,duplex} still
> remaining:
>
> if (MTK_HAS_CAPS(mac->hw->soc->caps,
> MTK_TRGMII_MT7621_CLK)) {
> ...
> } else {
> if (state->interface !=
> PHY_INTERFACE_MODE_TRGMII)
> mtk_gmac0_rgmii_adjust(mac->hw,
> state->speed);
>
> which also needs to be eliminated. Can that also be moved to
> mtk_mac_link_up()?
I know, you have pointed that out before. But I don't know how to fix
mtk_gmac0_rgmii_adjust(). This function changes the PLL of the MAC.
But without
documentation I am not sure what all the bits are used for.
Begin April I had a conversation with Sean about this. I also
explained what the
issue was. AFAIK he was going to take care of this issue.
Sean did you had time to resolve this issue?
Greats,
René
>
> Thanks.
>
>>
>> Thanks.
>>
>> drivers/net/ethernet/mediatek/mtk_eth_soc.c | 9 ++++-
>> drivers/net/ethernet/mediatek/mtk_eth_soc.h | 3 +-
>> drivers/net/ethernet/mediatek/mtk_sgmii.c | 37 +++++++++++++++------
>> 3 files changed, 36 insertions(+), 13 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> index 20db302d31ce..ef9ec3b6a5c8 100644
>> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> @@ -326,7 +326,7 @@ static void mtk_mac_config(struct
>> phylink_config *config, unsigned int mode,
>> /* Setup SGMIISYS with the determined property */
>> if (state->interface != PHY_INTERFACE_MODE_SGMII)
>> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
>> - state);
>> + state->interface);
>> else if (phylink_autoneg_inband(mode))
>> err = mtk_sgmii_setup_mode_an(eth->sgmii, sid);
>>
>> @@ -423,6 +423,13 @@ static void mtk_mac_link_up(struct
>> phylink_config *config,
>> phylink_config);
>> u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
>>
>> + if (phy_interface_mode_is_8023z(interface)) {
>> + /* Decide how GMAC and SGMIISYS be mapped */
>> + int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
>> + 0 : mac->id;
>> + mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
>> + }
>> +
>> mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
>> MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
>> MAC_MCR_FORCE_RX_FC);
>> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> index 454cfcd465fd..6f4b99bb7bfb 100644
>> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> @@ -932,7 +932,8 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct
>> device_node *np,
>> u32 ana_rgc3);
>> int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id);
>> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
>> - const struct phylink_link_state *state);
>> + phy_interface_t interface);
>> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed,
>> int duplex);
>> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id);
>>
>> int mtk_gmac_sgmii_path_setup(struct mtk_eth *eth, int mac_id);
>> diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> b/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> index 32d83421226a..372c85c830b5 100644
>> --- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> +++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> @@ -60,7 +60,7 @@ int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id)
>> }
>>
>> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
>> - const struct phylink_link_state *state)
>> + phy_interface_t interface)
>> {
>> unsigned int val;
>>
>> @@ -69,7 +69,7 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii
>> *ss, int id,
>>
>> regmap_read(ss->regmap[id], ss->ana_rgc3, &val);
>> val &= ~RG_PHY_SPEED_MASK;
>> - if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
>> + if (interface == PHY_INTERFACE_MODE_2500BASEX)
>> val |= RG_PHY_SPEED_3_125G;
>> regmap_write(ss->regmap[id], ss->ana_rgc3, val);
>>
>> @@ -78,11 +78,33 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii
>> *ss, int id,
>> val &= ~SGMII_AN_ENABLE;
>> regmap_write(ss->regmap[id], SGMSYS_PCS_CONTROL_1, val);
>>
>> + if (interface == PHY_INTERFACE_MODE_1000BASEX ||
>> + interface == PHY_INTERFACE_MODE_2500BASEX) {
>> + /* SGMII force mode setting */
>> + regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
>> + val &= ~SGMII_IF_MODE_MASK;
>> + val |= SGMII_SPEED_1000;
>> + val |= SGMII_DUPLEX_FULL;
>> + regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
>> + }
>> +
>> + /* Release PHYA power down state */
>> + regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
>> + val &= ~SGMII_PHYA_PWD;
>> + regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
>> +
>> + return 0;
>> +}
>> +
>> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex)
>> +{
>> + unsigned int val;
>> +
>> /* SGMII force mode setting */
>> regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
>> val &= ~SGMII_IF_MODE_MASK;
>>
>> - switch (state->speed) {
>> + switch (speed) {
>> case SPEED_10:
>> val |= SGMII_SPEED_10;
>> break;
>> @@ -95,17 +117,10 @@ int mtk_sgmii_setup_mode_force(struct
>> mtk_sgmii *ss, int id,
>> break;
>> }
>>
>> - if (state->duplex == DUPLEX_FULL)
>> + if (duplex == DUPLEX_FULL)
>> val |= SGMII_DUPLEX_FULL;
>>
>> regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
>> -
>> - /* Release PHYA power down state */
>> - regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
>> - val &= ~SGMII_PHYA_PWD;
>> - regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
>> -
>> - return 0;
>> }
>>
>> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id)
>> --
>> 2.20.1
>>
>>
>
> --
> RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
> FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
_______________________________________________
Linux-mediatek mailing list
Linux-mediatek@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-mediatek
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
@ 2020-06-30 22:13 ` René van Dorst
0 siblings, 0 replies; 13+ messages in thread
From: René van Dorst @ 2020-06-30 22:13 UTC (permalink / raw)
To: Russell King - ARM Linux admin, Sean Wang
Cc: Andrew Lunn, Florian Fainelli, netdev, Mark Lee, linux-mediatek,
John Crispin, Matthias Brugger, Jakub Kicinski, Heiner Kallweit,
David S. Miller, linux-arm-kernel, Felix Fietkau
Hi Russel and Sean,
Quoting Russell King - ARM Linux admin <linux@armlinux.org.uk>:
> On Tue, Jun 30, 2020 at 11:15:42AM +0100, Russell King wrote:
>> The SGMII PCS PHY needs to be updated with the link configuration in
>> the mac_link_up() call rather than in mac_config(). However,
>> mtk_sgmii_setup_mode_force() programs the SGMII block during
>> mac_config() when using 802.3z interface modes with the link
>> configuration.
>>
>> Split that functionality from mtk_sgmii_setup_mode_force(), moving it
>> to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
>>
>> This does not look correct to me: 802.3z modes operate at a fixed
>> speed. The contents of mtk_sgmii_link_up() look more appropriate for
>> SGMII mode, but the original code definitely did not call
>> mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
>>
>> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
>> ---
>> René, can you assist with this patch please - I really think there are
>> problems with the existing code. You call mtk_sgmii_setup_mode_force()
>> in a block which is conditionalised as:
>>
>> if (state->interface == PHY_INTERFACE_MODE_SGMII ||
>> phy_interface_mode_is_8023z(state->interface)) {
>> ...
>> if (state->interface != PHY_INTERFACE_MODE_SGMII)
>> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
>> state);
>>
>> Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
>> 2500BASE-X, which do not support anything but their native speeds.
>> Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
>> and 100M.
>>
>> Note that this patch is more about moving uses of state->{speed,duplex}
>> into mac_link_up(), rather than fixing this problem, but I don't think
>> the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
>> use.
>
> My Coccinelle script just found this use of state->{speed,duplex} still
> remaining:
>
> if (MTK_HAS_CAPS(mac->hw->soc->caps,
> MTK_TRGMII_MT7621_CLK)) {
> ...
> } else {
> if (state->interface !=
> PHY_INTERFACE_MODE_TRGMII)
> mtk_gmac0_rgmii_adjust(mac->hw,
> state->speed);
>
> which also needs to be eliminated. Can that also be moved to
> mtk_mac_link_up()?
I know, you have pointed that out before. But I don't know how to fix
mtk_gmac0_rgmii_adjust(). This function changes the PLL of the MAC.
But without
documentation I am not sure what all the bits are used for.
Begin April I had a conversation with Sean about this. I also
explained what the
issue was. AFAIK he was going to take care of this issue.
Sean did you had time to resolve this issue?
Greats,
René
>
> Thanks.
>
>>
>> Thanks.
>>
>> drivers/net/ethernet/mediatek/mtk_eth_soc.c | 9 ++++-
>> drivers/net/ethernet/mediatek/mtk_eth_soc.h | 3 +-
>> drivers/net/ethernet/mediatek/mtk_sgmii.c | 37 +++++++++++++++------
>> 3 files changed, 36 insertions(+), 13 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> index 20db302d31ce..ef9ec3b6a5c8 100644
>> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
>> @@ -326,7 +326,7 @@ static void mtk_mac_config(struct
>> phylink_config *config, unsigned int mode,
>> /* Setup SGMIISYS with the determined property */
>> if (state->interface != PHY_INTERFACE_MODE_SGMII)
>> err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
>> - state);
>> + state->interface);
>> else if (phylink_autoneg_inband(mode))
>> err = mtk_sgmii_setup_mode_an(eth->sgmii, sid);
>>
>> @@ -423,6 +423,13 @@ static void mtk_mac_link_up(struct
>> phylink_config *config,
>> phylink_config);
>> u32 mcr = mtk_r32(mac->hw, MTK_MAC_MCR(mac->id));
>>
>> + if (phy_interface_mode_is_8023z(interface)) {
>> + /* Decide how GMAC and SGMIISYS be mapped */
>> + int sid = (MTK_HAS_CAPS(eth->soc->caps, MTK_SHARED_SGMII)) ?
>> + 0 : mac->id;
>> + mtk_sgmii_link_up(eth->sgmii, sid, speed, duplex);
>> + }
>> +
>> mcr &= ~(MAC_MCR_SPEED_100 | MAC_MCR_SPEED_1000 |
>> MAC_MCR_FORCE_DPX | MAC_MCR_FORCE_TX_FC |
>> MAC_MCR_FORCE_RX_FC);
>> diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> index 454cfcd465fd..6f4b99bb7bfb 100644
>> --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
>> @@ -932,7 +932,8 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct
>> device_node *np,
>> u32 ana_rgc3);
>> int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id);
>> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
>> - const struct phylink_link_state *state);
>> + phy_interface_t interface);
>> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed,
>> int duplex);
>> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id);
>>
>> int mtk_gmac_sgmii_path_setup(struct mtk_eth *eth, int mac_id);
>> diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> b/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> index 32d83421226a..372c85c830b5 100644
>> --- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> +++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
>> @@ -60,7 +60,7 @@ int mtk_sgmii_setup_mode_an(struct mtk_sgmii *ss, int id)
>> }
>>
>> int mtk_sgmii_setup_mode_force(struct mtk_sgmii *ss, int id,
>> - const struct phylink_link_state *state)
>> + phy_interface_t interface)
>> {
>> unsigned int val;
>>
>> @@ -69,7 +69,7 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii
>> *ss, int id,
>>
>> regmap_read(ss->regmap[id], ss->ana_rgc3, &val);
>> val &= ~RG_PHY_SPEED_MASK;
>> - if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
>> + if (interface == PHY_INTERFACE_MODE_2500BASEX)
>> val |= RG_PHY_SPEED_3_125G;
>> regmap_write(ss->regmap[id], ss->ana_rgc3, val);
>>
>> @@ -78,11 +78,33 @@ int mtk_sgmii_setup_mode_force(struct mtk_sgmii
>> *ss, int id,
>> val &= ~SGMII_AN_ENABLE;
>> regmap_write(ss->regmap[id], SGMSYS_PCS_CONTROL_1, val);
>>
>> + if (interface == PHY_INTERFACE_MODE_1000BASEX ||
>> + interface == PHY_INTERFACE_MODE_2500BASEX) {
>> + /* SGMII force mode setting */
>> + regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
>> + val &= ~SGMII_IF_MODE_MASK;
>> + val |= SGMII_SPEED_1000;
>> + val |= SGMII_DUPLEX_FULL;
>> + regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
>> + }
>> +
>> + /* Release PHYA power down state */
>> + regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
>> + val &= ~SGMII_PHYA_PWD;
>> + regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
>> +
>> + return 0;
>> +}
>> +
>> +void mtk_sgmii_link_up(struct mtk_sgmii *ss, int id, int speed, int duplex)
>> +{
>> + unsigned int val;
>> +
>> /* SGMII force mode setting */
>> regmap_read(ss->regmap[id], SGMSYS_SGMII_MODE, &val);
>> val &= ~SGMII_IF_MODE_MASK;
>>
>> - switch (state->speed) {
>> + switch (speed) {
>> case SPEED_10:
>> val |= SGMII_SPEED_10;
>> break;
>> @@ -95,17 +117,10 @@ int mtk_sgmii_setup_mode_force(struct
>> mtk_sgmii *ss, int id,
>> break;
>> }
>>
>> - if (state->duplex == DUPLEX_FULL)
>> + if (duplex == DUPLEX_FULL)
>> val |= SGMII_DUPLEX_FULL;
>>
>> regmap_write(ss->regmap[id], SGMSYS_SGMII_MODE, val);
>> -
>> - /* Release PHYA power down state */
>> - regmap_read(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, &val);
>> - val &= ~SGMII_PHYA_PWD;
>> - regmap_write(ss->regmap[id], SGMSYS_QPHY_PWR_STATE_CTRL, val);
>> -
>> - return 0;
>> }
>>
>> void mtk_sgmii_restart_an(struct mtk_eth *eth, int mac_id)
>> --
>> 2.20.1
>>
>>
>
> --
> RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
> FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
2020-06-30 22:13 ` René van Dorst
(?)
@ 2020-07-01 0:05 ` Russell King - ARM Linux admin
-1 siblings, 0 replies; 13+ messages in thread
From: Russell King - ARM Linux admin @ 2020-07-01 0:05 UTC (permalink / raw)
To: René van Dorst
Cc: Sean Wang, Andrew Lunn, Florian Fainelli, Heiner Kallweit,
David S. Miller, netdev, Felix Fietkau, John Crispin, Mark Lee,
Jakub Kicinski, Matthias Brugger, linux-arm-kernel,
linux-mediatek
On Tue, Jun 30, 2020 at 10:13:08PM +0000, René van Dorst wrote:
> Hi Russel and Sean,
>
> Quoting Russell King - ARM Linux admin <linux@armlinux.org.uk>:
>
> > On Tue, Jun 30, 2020 at 11:15:42AM +0100, Russell King wrote:
> > > The SGMII PCS PHY needs to be updated with the link configuration in
> > > the mac_link_up() call rather than in mac_config(). However,
> > > mtk_sgmii_setup_mode_force() programs the SGMII block during
> > > mac_config() when using 802.3z interface modes with the link
> > > configuration.
> > >
> > > Split that functionality from mtk_sgmii_setup_mode_force(), moving it
> > > to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
> > >
> > > This does not look correct to me: 802.3z modes operate at a fixed
> > > speed. The contents of mtk_sgmii_link_up() look more appropriate for
> > > SGMII mode, but the original code definitely did not call
> > > mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
> > >
> > > Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
> > > ---
> > > René, can you assist with this patch please - I really think there are
> > > problems with the existing code. You call mtk_sgmii_setup_mode_force()
> > > in a block which is conditionalised as:
> > >
> > > if (state->interface == PHY_INTERFACE_MODE_SGMII ||
> > > phy_interface_mode_is_8023z(state->interface)) {
> > > ...
> > > if (state->interface != PHY_INTERFACE_MODE_SGMII)
> > > err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
> > > state);
> > >
> > > Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
> > > 2500BASE-X, which do not support anything but their native speeds.
> > > Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
> > > and 100M.
> > >
> > > Note that this patch is more about moving uses of state->{speed,duplex}
> > > into mac_link_up(), rather than fixing this problem, but I don't think
> > > the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
> > > use.
> >
> > My Coccinelle script just found this use of state->{speed,duplex} still
> > remaining:
> >
> > if (MTK_HAS_CAPS(mac->hw->soc->caps,
> > MTK_TRGMII_MT7621_CLK)) {
> > ...
> > } else {
> > if (state->interface !=
> > PHY_INTERFACE_MODE_TRGMII)
> > mtk_gmac0_rgmii_adjust(mac->hw,
> > state->speed);
> >
> > which also needs to be eliminated. Can that also be moved to
> > mtk_mac_link_up()?
>
> I know, you have pointed that out before. But I don't know how to fix
> mtk_gmac0_rgmii_adjust(). This function changes the PLL of the MAC. But
> without documentation I am not sure what all the bits are used for.
I'd forgotten...
> Begin April I had a conversation with Sean about this. I also explained what
> the issue was. AFAIK he was going to take care of this issue.
>
> Sean did you had time to resolve this issue?
Well, I think the code as it stands is quite broken.
If we start a bit earlier in mtk_mac_config(), we have this:
if (!MTK_HAS_CAPS(eth->soc->caps, MTK_SOC_MT7628) &&
mac->interface != state->interface) {
which prevents us entering this block unless the interface mode has
changed and we are not MT7628. This block of code includes the two
calls to mtk_gmac0_rgmii_adjust(), which are dependent on
state->speed.
Since mac->interface starts off as PHY_INTERFACE_MODE_NA, the first
time we head into mtk_mac_config(), the interface mode will be
different, and we will enter this block of code, maybe calling down
into mtk_gmac0_rgmii_adjust() if appropriate.
The first call will be via phylink_start(), which will call it with
the initial configuration - the link will be down, and state->speed
will be SPEED_UNKNOWN. So, the various tests inside
mtk_gmac0_rgmii_adjust() for speed == SPEED_1000 will all be false,
meaning it'll program it as if for 10M or 100M speeds.
When the link comes up, yes, mtk_mac_config() will be called again
with the link parameters, but state->interface will now match
mac->interface - so the block of code containing the call to
mtk_gmac0_rgmii_adjust() will not be entered, and so none of that
code gets executed when the link comes up/down.
Now, if I dig out object 8ddbb8dcf032 from the git repository, which
was the state of the file immedately prior to the phylink conversion,
I find:
static void mtk_phy_link_adjust(struct net_device *dev)
{
This is the function that phylib would call when the link comes up
or down. It tests for MTK_RESETTING, starts preparing a value for
mcr, and then:
if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_GMAC1_TRGMII) && !mac->id) {
if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_TRGMII_MT7621_CLK)) {
if (mt7621_gmac0_rgmii_adjust(mac->hw,
dev->phydev->interface))
return;
} else {
if (!mac->trgmii)
mtk_gmac0_rgmii_adjust(mac->hw,
dev->phydev->speed);
}
}
It then finishes creating a value for mcr, before writing it to the
register, and printing the link status to the kernel log.
Hence, mtk_gmac0_rgmii_adjust() would've been called every time there's
a change of link state, and is expected to be passed the current speed.
There seems to be a difference in behaviour between the pre-phylink and
post-phylink drivers, and I think moving mtk_gmac0_rgmii_adjust() into
mtk_mac_link_up() would be a definite improvement, possibly even a
regression fix.
However, it would be reasonable to assume that there should be reports
that mtk_eth_soc doesn't work if this were the case. So... odd.
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
@ 2020-07-01 0:05 ` Russell King - ARM Linux admin
0 siblings, 0 replies; 13+ messages in thread
From: Russell King - ARM Linux admin @ 2020-07-01 0:05 UTC (permalink / raw)
To: René van Dorst
Cc: Andrew Lunn, Florian Fainelli, netdev, Sean Wang, Mark Lee,
linux-mediatek, John Crispin, Matthias Brugger, Jakub Kicinski,
Felix Fietkau, David S. Miller, linux-arm-kernel,
Heiner Kallweit
On Tue, Jun 30, 2020 at 10:13:08PM +0000, René van Dorst wrote:
> Hi Russel and Sean,
>
> Quoting Russell King - ARM Linux admin <linux@armlinux.org.uk>:
>
> > On Tue, Jun 30, 2020 at 11:15:42AM +0100, Russell King wrote:
> > > The SGMII PCS PHY needs to be updated with the link configuration in
> > > the mac_link_up() call rather than in mac_config(). However,
> > > mtk_sgmii_setup_mode_force() programs the SGMII block during
> > > mac_config() when using 802.3z interface modes with the link
> > > configuration.
> > >
> > > Split that functionality from mtk_sgmii_setup_mode_force(), moving it
> > > to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
> > >
> > > This does not look correct to me: 802.3z modes operate at a fixed
> > > speed. The contents of mtk_sgmii_link_up() look more appropriate for
> > > SGMII mode, but the original code definitely did not call
> > > mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
> > >
> > > Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
> > > ---
> > > René, can you assist with this patch please - I really think there are
> > > problems with the existing code. You call mtk_sgmii_setup_mode_force()
> > > in a block which is conditionalised as:
> > >
> > > if (state->interface == PHY_INTERFACE_MODE_SGMII ||
> > > phy_interface_mode_is_8023z(state->interface)) {
> > > ...
> > > if (state->interface != PHY_INTERFACE_MODE_SGMII)
> > > err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
> > > state);
> > >
> > > Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
> > > 2500BASE-X, which do not support anything but their native speeds.
> > > Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
> > > and 100M.
> > >
> > > Note that this patch is more about moving uses of state->{speed,duplex}
> > > into mac_link_up(), rather than fixing this problem, but I don't think
> > > the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
> > > use.
> >
> > My Coccinelle script just found this use of state->{speed,duplex} still
> > remaining:
> >
> > if (MTK_HAS_CAPS(mac->hw->soc->caps,
> > MTK_TRGMII_MT7621_CLK)) {
> > ...
> > } else {
> > if (state->interface !=
> > PHY_INTERFACE_MODE_TRGMII)
> > mtk_gmac0_rgmii_adjust(mac->hw,
> > state->speed);
> >
> > which also needs to be eliminated. Can that also be moved to
> > mtk_mac_link_up()?
>
> I know, you have pointed that out before. But I don't know how to fix
> mtk_gmac0_rgmii_adjust(). This function changes the PLL of the MAC. But
> without documentation I am not sure what all the bits are used for.
I'd forgotten...
> Begin April I had a conversation with Sean about this. I also explained what
> the issue was. AFAIK he was going to take care of this issue.
>
> Sean did you had time to resolve this issue?
Well, I think the code as it stands is quite broken.
If we start a bit earlier in mtk_mac_config(), we have this:
if (!MTK_HAS_CAPS(eth->soc->caps, MTK_SOC_MT7628) &&
mac->interface != state->interface) {
which prevents us entering this block unless the interface mode has
changed and we are not MT7628. This block of code includes the two
calls to mtk_gmac0_rgmii_adjust(), which are dependent on
state->speed.
Since mac->interface starts off as PHY_INTERFACE_MODE_NA, the first
time we head into mtk_mac_config(), the interface mode will be
different, and we will enter this block of code, maybe calling down
into mtk_gmac0_rgmii_adjust() if appropriate.
The first call will be via phylink_start(), which will call it with
the initial configuration - the link will be down, and state->speed
will be SPEED_UNKNOWN. So, the various tests inside
mtk_gmac0_rgmii_adjust() for speed == SPEED_1000 will all be false,
meaning it'll program it as if for 10M or 100M speeds.
When the link comes up, yes, mtk_mac_config() will be called again
with the link parameters, but state->interface will now match
mac->interface - so the block of code containing the call to
mtk_gmac0_rgmii_adjust() will not be entered, and so none of that
code gets executed when the link comes up/down.
Now, if I dig out object 8ddbb8dcf032 from the git repository, which
was the state of the file immedately prior to the phylink conversion,
I find:
static void mtk_phy_link_adjust(struct net_device *dev)
{
This is the function that phylib would call when the link comes up
or down. It tests for MTK_RESETTING, starts preparing a value for
mcr, and then:
if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_GMAC1_TRGMII) && !mac->id) {
if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_TRGMII_MT7621_CLK)) {
if (mt7621_gmac0_rgmii_adjust(mac->hw,
dev->phydev->interface))
return;
} else {
if (!mac->trgmii)
mtk_gmac0_rgmii_adjust(mac->hw,
dev->phydev->speed);
}
}
It then finishes creating a value for mcr, before writing it to the
register, and printing the link status to the kernel log.
Hence, mtk_gmac0_rgmii_adjust() would've been called every time there's
a change of link state, and is expected to be passed the current speed.
There seems to be a difference in behaviour between the pre-phylink and
post-phylink drivers, and I think moving mtk_gmac0_rgmii_adjust() into
mtk_mac_link_up() would be a definite improvement, possibly even a
regression fix.
However, it would be reasonable to assume that there should be reports
that mtk_eth_soc doesn't work if this were the case. So... odd.
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
_______________________________________________
Linux-mediatek mailing list
Linux-mediatek@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-mediatek
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY
@ 2020-07-01 0:05 ` Russell King - ARM Linux admin
0 siblings, 0 replies; 13+ messages in thread
From: Russell King - ARM Linux admin @ 2020-07-01 0:05 UTC (permalink / raw)
To: René van Dorst
Cc: Andrew Lunn, Florian Fainelli, netdev, Sean Wang, Mark Lee,
linux-mediatek, John Crispin, Matthias Brugger, Jakub Kicinski,
Felix Fietkau, David S. Miller, linux-arm-kernel,
Heiner Kallweit
On Tue, Jun 30, 2020 at 10:13:08PM +0000, René van Dorst wrote:
> Hi Russel and Sean,
>
> Quoting Russell King - ARM Linux admin <linux@armlinux.org.uk>:
>
> > On Tue, Jun 30, 2020 at 11:15:42AM +0100, Russell King wrote:
> > > The SGMII PCS PHY needs to be updated with the link configuration in
> > > the mac_link_up() call rather than in mac_config(). However,
> > > mtk_sgmii_setup_mode_force() programs the SGMII block during
> > > mac_config() when using 802.3z interface modes with the link
> > > configuration.
> > >
> > > Split that functionality from mtk_sgmii_setup_mode_force(), moving it
> > > to a new mtk_sgmii_link_up() function, and call it from mac_link_up().
> > >
> > > This does not look correct to me: 802.3z modes operate at a fixed
> > > speed. The contents of mtk_sgmii_link_up() look more appropriate for
> > > SGMII mode, but the original code definitely did not call
> > > mtk_sgmii_setup_mode_force() for SGMII mode but only 802.3z mode.
> > >
> > > Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
> > > ---
> > > René, can you assist with this patch please - I really think there are
> > > problems with the existing code. You call mtk_sgmii_setup_mode_force()
> > > in a block which is conditionalised as:
> > >
> > > if (state->interface == PHY_INTERFACE_MODE_SGMII ||
> > > phy_interface_mode_is_8023z(state->interface)) {
> > > ...
> > > if (state->interface != PHY_INTERFACE_MODE_SGMII)
> > > err = mtk_sgmii_setup_mode_force(eth->sgmii, sid,
> > > state);
> > >
> > > Hence, mtk_sgmii_setup_mode_force() is only called for 1000BASE-X and
> > > 2500BASE-X, which do not support anything but their native speeds.
> > > Yet, mtk_sgmii_setup_mode_force() tries to program the SGMII for 10M
> > > and 100M.
> > >
> > > Note that this patch is more about moving uses of state->{speed,duplex}
> > > into mac_link_up(), rather than fixing this problem, but I don't think
> > > the addition in mtk_mac_link_up(), nor mtk_sgmii_link_up() is of any
> > > use.
> >
> > My Coccinelle script just found this use of state->{speed,duplex} still
> > remaining:
> >
> > if (MTK_HAS_CAPS(mac->hw->soc->caps,
> > MTK_TRGMII_MT7621_CLK)) {
> > ...
> > } else {
> > if (state->interface !=
> > PHY_INTERFACE_MODE_TRGMII)
> > mtk_gmac0_rgmii_adjust(mac->hw,
> > state->speed);
> >
> > which also needs to be eliminated. Can that also be moved to
> > mtk_mac_link_up()?
>
> I know, you have pointed that out before. But I don't know how to fix
> mtk_gmac0_rgmii_adjust(). This function changes the PLL of the MAC. But
> without documentation I am not sure what all the bits are used for.
I'd forgotten...
> Begin April I had a conversation with Sean about this. I also explained what
> the issue was. AFAIK he was going to take care of this issue.
>
> Sean did you had time to resolve this issue?
Well, I think the code as it stands is quite broken.
If we start a bit earlier in mtk_mac_config(), we have this:
if (!MTK_HAS_CAPS(eth->soc->caps, MTK_SOC_MT7628) &&
mac->interface != state->interface) {
which prevents us entering this block unless the interface mode has
changed and we are not MT7628. This block of code includes the two
calls to mtk_gmac0_rgmii_adjust(), which are dependent on
state->speed.
Since mac->interface starts off as PHY_INTERFACE_MODE_NA, the first
time we head into mtk_mac_config(), the interface mode will be
different, and we will enter this block of code, maybe calling down
into mtk_gmac0_rgmii_adjust() if appropriate.
The first call will be via phylink_start(), which will call it with
the initial configuration - the link will be down, and state->speed
will be SPEED_UNKNOWN. So, the various tests inside
mtk_gmac0_rgmii_adjust() for speed == SPEED_1000 will all be false,
meaning it'll program it as if for 10M or 100M speeds.
When the link comes up, yes, mtk_mac_config() will be called again
with the link parameters, but state->interface will now match
mac->interface - so the block of code containing the call to
mtk_gmac0_rgmii_adjust() will not be entered, and so none of that
code gets executed when the link comes up/down.
Now, if I dig out object 8ddbb8dcf032 from the git repository, which
was the state of the file immedately prior to the phylink conversion,
I find:
static void mtk_phy_link_adjust(struct net_device *dev)
{
This is the function that phylib would call when the link comes up
or down. It tests for MTK_RESETTING, starts preparing a value for
mcr, and then:
if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_GMAC1_TRGMII) && !mac->id) {
if (MTK_HAS_CAPS(mac->hw->soc->caps, MTK_TRGMII_MT7621_CLK)) {
if (mt7621_gmac0_rgmii_adjust(mac->hw,
dev->phydev->interface))
return;
} else {
if (!mac->trgmii)
mtk_gmac0_rgmii_adjust(mac->hw,
dev->phydev->speed);
}
}
It then finishes creating a value for mcr, before writing it to the
register, and printing the link status to the kernel log.
Hence, mtk_gmac0_rgmii_adjust() would've been called every time there's
a change of link state, and is expected to be passed the current speed.
There seems to be a difference in behaviour between the pre-phylink and
post-phylink drivers, and I think moving mtk_gmac0_rgmii_adjust() into
mtk_mac_link_up() would be a definite improvement, possibly even a
regression fix.
However, it would be reasonable to assume that there should be reports
that mtk_eth_soc doesn't work if this were the case. So... odd.
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 13+ messages in thread
end of thread, other threads:[~2020-07-01 0:07 UTC | newest]
Thread overview: 13+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-06-30 10:15 [PATCH RFC net-next] net: mtk_eth_soc: use resolved link config for PCS PHY Russell King
2020-06-30 10:15 ` Russell King
2020-06-30 10:15 ` Russell King
2020-06-30 10:46 ` Russell King - ARM Linux admin
2020-06-30 10:46 ` Russell King - ARM Linux admin
2020-06-30 10:46 ` Russell King - ARM Linux admin
2020-06-30 22:13 ` René van Dorst
2020-06-30 22:13 ` René van Dorst
2020-06-30 22:13 ` René van Dorst
2020-07-01 0:05 ` Russell King - ARM Linux admin
2020-07-01 0:05 ` Russell King - ARM Linux admin
2020-07-01 0:05 ` Russell King - ARM Linux admin
2020-06-30 15:29 ` kernel test robot
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.